Contains the performance analysis of various deep learning frameworks such as Tensorflow, Pytorch, Scikit Learn. A CNN trained on fashion MNIST written in all the above frameworks is used for comparison. Done as a part of Performance Modelling Project- IT 364

Steps to obtain heap profiling

• Install valgrind software
• Save .ipynb as .py file
• run the command valgrind --tool=massif python3 example.py
• Note the process id while it executes ( execution is very slow )
• Heap profile statistics will be stored in massif.out.